Strava gets an AI running coach

A new integration for the fitness-tracking app Strava allows subscribers to query workout data using natural language and receive detailed training insights. However, not within the app itself.

The feature connects Strava to Claude, Anthropic’s large language model (LLM), through a Model Context Protocol (MCP) connector. This gives Claude read-only access to a subscriber’s Strava data, allowing the AI assistant to analyse training history and answer personalised questions about performance, recovery and progress.

“Until now, anyone who wanted to analyse their Strava history with an AI assistant had to follow a complicated and multi-step process that required an export of their data, a subsequent copy and paste of that data into an LLM by hand every separate time they wanted to receive an update, or rely on unsecure third-party tools,” says Strava. “The new MCP makes that process much easier by providing a seamless, secure experience connecting Claude to a breadth of activity data.”

The company says these include: 

• Access to full stream data – per second heart rate, and pace.
• GPS data for geographic analysis of activities.
• Power data for cycling.
• Access to your club and event data.

Ryan Dixon, Strava VP for partnerships and developer relations, says: “Athletes have been telling us, in increasingly creative ways, that they want more ways to analyse their own training data. They’ve been doing it with spreadsheets, exports, and third-party scripts for years. The MCP connector gives them a far more efficient, safer tool while keeping the athlete in control, representing a step-change in our subscription offering.”

The launch came days before Strava and Runna’s Global Running Day challenge on 3 June 2026, which aimed to set a record for the most 5km activities completed in a single day. Qualifying Strava modes included hike, run, trail run, walk, and wheelchair.

About 850,000 participants collectively covered roughly 2.5-million miles (about 4-million kilometres), equivalent to approximately 100 laps around the Earth. Events such as these generate the training data that subscribers can now analyse through the Claude integration.

According to Strava, the AI assistant can answer questions like:

• “What types of activities have most improved my fitness?” – Unlock your full training history, analysed in plain language. Volume trends, pace over time, how specific training cycles have stacked up against your best.
• “Are my easy days easy enough?” – It’s hard to keep your pace low – look back at all your easier efforts in context to inform your perspective.
• “How is my cross-training affecting my running?” – Synthesise trends across strength, cycling, running, or however you stay active.

The MCP connector provides read-only access to Strava data, is limited to a subscriber’s account, and can be revoked through the app’s settings.

* To set up the Strava MCP connector, follow the instructions on the Strava’s website here.